The NCBO creates tools and services that provide the ontological backbone for a broad range of scientific disciplines. We are producing a scalable, distributed infrastructure for publishing, managing, and evaluating ontologies. We are also creating ontology-based annotations of biomedical data sources to advance biomedical, clinical, and translational research. In the next five years, we will continue to grow the BioPortal library and services that will enable scientists to custom-tailor library components and to embed these components directly in their applications. The first generation of BioPortal focused on the content and services for the bio-ontology community. The next generation will provide additional content and services for the broader community of biomedical, clinical, and translational researchers, with an emphasis on generalizability and wide-scale adoption. In many cases, the ontology needs of a research area evolve quickly and researchers must be able to add new ontology terms and to refine and extend existing ones in order to meet community and institutional requirements. Our goal is to enable "agile ontology development"[unreadable]the incremental, user-guided, needs-driven evolution of ontologies. To accomplish this goal, we will integrate the infrastructure for ontology publishing, review, and application with the ontology-revision process, to create a common infrastructure for user feedback, change proposals, and updates by developers. We will use the annotation tools that we have already developed to annotate automatically a large set of public biomedical resources, creating a comprehensive index of ontology-based annotations for enabling translational discoveries. Finally, we will develop analytic methods to profile biomedical data sets for enrichment against the background of these annotations. As a result, just as biologists use the Gene Ontology to determine biological processes over-represented (or enriched) in a set of differentially expressed genes, clinical and translational investigators will be able to determine enrichment of terms representing diseases (or class of diseases), drugs, or other controlled terms in data sets of their choice.